PulseAugur
EN
LIVE 07:54:54
(CA) Adaptive directional gradients for parameterised quantum circuits

New quantum gradient method slashes training costs

Researchers have developed a new framework for estimating gradients in parameterised quantum circuits, aiming to reduce the high measurement costs associated with training. This method, based on the forward mode of automatic differentiation, uses random directional derivatives to provide an unbiased gradient estimator. The proposed QUIVER optimizer, derived from this framework, demonstrates significant efficiency gains in training quantum neural networks and outperforms existing methods on various optimization problems. AI

IMPACT Reduces computational overhead for training quantum models, potentially accelerating research and development in quantum machine learning.

RANK_REASON Academic paper detailing a new method for training quantum circuits. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →

COVERAGE [2]

  1. arXiv cs.LG TIER_1 (CA) · Brian Coyle, Snehal Raj, Virag Umathe, El Amine Cherrat, Elham Kashefi ·

    Adaptive directional gradients for parameterized quantum circuits

    arXiv:2606.09734v1 Announce Type: cross Abstract: Training parameterised quantum circuits (PQCs) on quantum hardware is bottlenecked by the measurement cost of gradient estimation, which under the parameter-shift rule scales linearly in the number of trainable parameters and domi…

  2. arXiv cs.LG TIER_1 (CA) · Elham Kashefi ·

    Adaptive directional gradients for parameterized quantum circuits

    Training parameterised quantum circuits (PQCs) on quantum hardware is bottlenecked by the measurement cost of gradient estimation, which under the parameter-shift rule scales linearly in the number of trainable parameters and dominates the total shot budget of training at scale. …